What are the serum biomarkers for dementia and its preclinical phase?

Serum Biomarkers for Dementia and Preclinical Disease

The most clinically validated serum biomarkers for Alzheimer's disease and its preclinical phase are plasma phosphorylated tau (p-tau217, p-tau181, p-tau231), plasma Aβ42/Aβ40 ratio, glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL), with plasma p-tau showing the strongest performance for detecting AD pathology across all disease stages. 1

Core Blood-Based Biomarkers

Plasma Phosphorylated Tau (p-tau)

• Plasma p-tau (particularly p-tau217) is the most promising blood biomarker for symptomatic AD, demonstrating areas under the curve (AUC) of 0.92-0.98 for predicting amyloid status and showing excellent performance during both preclinical and symptomatic disease stages 2, 3
• P-tau181, p-tau217, and p-tau231 isoforms all demonstrate desirable sensitivity and specificity for identifying individuals with AD pathology 3
• Plasma p-tau181 achieves an AUC of 0.78-0.81 for discriminating A+T+ from A-T- participants, with moderate correlation to CSF levels (Rho=0.51) 4
• P-tau biomarkers may be combined with APOE genotype or Aβ42/Aβ40 during prodromal disease stages for enhanced accuracy 1

Plasma Amyloid Beta (Aβ42/Aβ40 Ratio)

• Plasma Aβ42/Aβ40 ratio reflects selective depletion of Aβ42 due to cerebral plaque deposition, similar to CSF patterns 1
• The major limitation is the small fold change between Aβ-positive and Aβ-negative individuals (only 8-15% reduction compared to 40-60% in CSF), making it less robust than p-tau markers 1
• Immunoprecipitation mass spectrometry (IP-MS) methods achieve AUC of 0.86 for detecting Aβ pathology, outperforming immunoassays (AUC 0.69-0.78) 1
• Plasma Aβ composite shows AUC of 0.75 for discriminating positive and negative participants, with moderate correlation to CSF Aβ1-42/Aβ1-40 (Rho=-0.5) 4
• Levels are fully changed during pre-symptomatic disease stages, allowing identification of Aβ pathology in cognitively unimpaired individuals with accuracies comparable to symptomatic patients 1

Neurofilament Light Chain (NfL)

• Plasma NfL is a non-specific marker of neuronal injury and neurodegeneration, not exclusive to AD but useful across multiple neurodegenerative diseases 1, 5, 3
• NfL demonstrates high correlation between plasma and CSF levels (Rho=0.78), with AUC of 0.88 for discriminating neurodegeneration categories 4
• Can detect onset of neurodegeneration in mutation carriers entering the clinical disease phase 1
• Useful for monitoring disease-modifying treatment effects, as demonstrated in multiple sclerosis, spinal muscular atrophy, and HIV-associated neurocognitive dysfunction 1

Glial Fibrillary Acidic Protein (GFAP)

• GFAP is a marker of neuro-immune activation and astrocytic response, showing promise for AD identification 1, 3
• Demonstrates desirable sensitivity and specificity for identifying individuals with AD pathology and dementia 3

Biomarker Staging Framework

Preclinical AD Classification

The biomarker model follows a temporal sequence where markers become abnormal in ordered stages 1:

Stage 1: Amyloid positivity alone

• Detectable via reduced plasma Aβ42/Aβ40 ratio
• Represents earliest detectable pathological change
• May precede clinical symptoms by more than a decade 1

Stage 2: Amyloid positivity + neurodegeneration markers

• Elevated plasma p-tau (reflecting neuronal injury)
• Elevated NfL (reflecting axonal damage)
• Individuals at this stage are farther along the disease trajectory 1

Stage 3: Amyloid + neurodegeneration + subtle cognitive decline

• All biomarkers abnormal plus emerging cognitive changes
• Approaching the border with mild cognitive impairment 1

Clinical Application Guidelines

Current Appropriate Use

Blood biomarkers with established thresholds (especially plasma Aβ42/Aβ40 and p-tau) can be used as first-line screening tools, but AD status must be confirmed with PET or CSF before clinical decisions or trial enrollment 1

Key Limitations and Caveats

• Blood biomarkers currently require confirmation and should not be used as stand-alone diagnostic tools in clinical practice, as positive predictive values need to exceed 90-95% for definitive diagnosis 1
• The combination of plasma Aβ composite, pTau181, and NfL does not outperform pTau181 alone for discriminating A+T+ from A-T- participants 4
• Complement C3 and alpha-2-macroglobulin, despite initial promise, have not demonstrated sufficient predictive value for clinical use 6

Clinical Trial Applications

• Blood biomarkers substantially reduce screening costs and time for trial enrollment, as demonstrated by the A4 trial requiring 3.5 years and >4000 amyloid PET scans to enroll 1169 participants 1
• In non-AD trials, plasma Aβ42/Aβ40 and p-tau can exclude patients with AD co-pathology 1
• Plasma NfL is being used to screen for genetic neurodegenerative diseases like frontotemporal dementia with GRN or MAPT mutations 1

Practical Algorithm for Biomarker Use

For symptomatic patients (MCI or dementia):

1. Measure plasma p-tau (preferably p-tau217) as primary screening marker 1, 3
2. Add plasma Aβ42/Aβ40 if p-tau results are equivocal 1
3. Confirm positive results with CSF or amyloid PET before treatment decisions 1

For preclinical/asymptomatic individuals:

1. Combine plasma Aβ42/Aβ40 with p-tau for optimal detection 1
2. Consider APOE genotype as adjunctive risk stratification 1
3. Mandatory confirmation with PET or CSF required before any intervention 1
4. Treatment is inappropriate for cognitively unimpaired individuals even with positive biomarkers 2, 7

For monitoring neurodegeneration:

• Use plasma NfL as a general marker of neuronal injury across disease stages 1, 5